Munich Personal RePEc Archive
Decades not lost, but won: increased employment, higher wages, and more equal opportunities in the Japanese labour market
Blind, Georg and Lottanti von Mandach, Stefania
University of Zurich
September 2014
Online at https://mpra.ub.uni-muenchen.de/66845/
MPRA Paper No. 66845, posted 24 Sep 2015 08:42 UTC
Decades not lost, but won: increased employment, higher wages, and more equal opportunities in the Japanese labour market
Georg Blind and Stefania Lottanti von Mandach1
Abstract
We take a fresh analytical look at the developments in the Japanese labour market over the last two decades and provide answers to three key questions: first, was regular employment replaced by non-regular employment on an aggregate level in general and at the expense of younger generations in particular? Second, how do today’s labour market outcomes compare to the heydays of the Japanese economy, i.e. to the late 1980s? Third, have wage gaps between men and women as well as between regular and non-regular employees increased, stayed the same, or shrunk? Contrary to both public perception and many research opinions, we find that the Japanese labour market as a whole has produced normatively positive outcomes of an unexpected magnitude in a long-term perspective. By 2010, employment has expanded beyond the structural frame of 1988 both in its non-standard and in its standard segment.We further find that the increase in non-regular employment was predominantly due to an increase in labour demand, mirrored by a significantly narrowed wage gap between non-standard and standard employment. Lastly, we find that women have made substantial inroads into the labour market, both in terms of regular employment and real wage development.
Keywords: economics of gender; labour force and employment; wage differentials JEL classifications: J16; J21; J31
1. Introduction
Wage inequality in advanced economies has increased in recent years and has become an important issue in public discourse due its socio-economic impact (Antonczyk et al. 2010; Chusseau and Dumont 2012). One dimension of these developments is the perceived decline of standard or regular full-time employment relative to the growth of non-regular or atypical employment, i.e., positions that are either not full-time and/or not permanent, and therefore generally less well paid, less stable and less likely to include social security benefits (Beck 1992; Bernardt and Marcotte 2000; Appelbaum 2002; Diekmann and Jann 2005;
Hevenstone 2008; OECD 2009; JILPT 2011). A second dimension is the hotly debated gender gap. Where both dimensions increasingly intersect, the compound effect on women in non-regular employment is seen as epitomizing a trend to greater wage inequality. A third dimension is inter-generational inequality, with increasing numbers of young adults unable to make their way into standard employment.
1 The authors are grateful to the editor for encouraging them to further develop this article and to the anonymous
Japan is no exception to the trend toward more non-regular employment. The OECD (2009: 87) observes a
“particularly strong expansion of non-standard forms of employment.” According to the Japanese Ministry of Health, Labour, and Welfare (MHLW), the share of non-standard employment (part-time, temporary, and contract employees, as well as day labourers) has risen from 15.4% to 30.0% between 1985 and 2010 (MHLW 2012a). Observers label this development “the most important change in the Japanese labour market” (Keizer 2008: 411). With 2013 figures indicating that Japan has the second largest gender wage gap in the OECD, concerns abound regarding the development of the Japanese labour market over the last two decades.2
In light of this development, academic assessments of labour market outcomes in Japan have been
overwhelmingly negative. The de-standardization of work is blamed for creating economic and social gaps between groups of employees, since individuals in non-standard employment face less favourable conditions in terms of social security, hourly wages, bonuses, retirement benefits, and contract period (Yamagami 2006;
Weathers 2009; Fu 2011). According to Song (2010), the rapidly expanded practice of hiring non-regular employees was fostered through a number of regulatory changes in the Japanese labour market during the 1990s and 2000s.
With regard to the second dimension, Shire and Imai (2000), Weathers (2001), Macnaughtan (2006), and Yamada (2009), point to a continuing and intensifying gender bias in the labour market. Although female labour market participation is increasing, the high percentage of non-standard female employees throughout all age groups further consolidates the inferior position of women in the Japanese labour market and in the social welfare system. For Katō and Kambayashi (2013), Japanese women have lost what they gained during the decade preceding the economic stagnation of the 1990s, namely the inroads made into the primary segment of the labour market. Between 1982 and 1992, the proportion of female standard employees on indefinite contracts increased from 23.8% to 26.6%, but fell to 22.1% between 1992 and 2002.
As for intergenerational inequality, Imai (2004) finds that replacing standard with non-standard workers is taking place as “an inter-generational project in which older standard workers are replaced with younger non-standard workers”. Ishiguro (2008) points out that the ratio of non-standard employment is especially high for the cohorts aged 15 to 19 and 20 to 24 and warns that, given the characteristics of the Japanese employment system, the chances that younger generations in non-standard employment will upgrade into standard employment are low.
In contrast, very few studies have identified normatively positive evidence on the phenomenon of atypical employment. For instance, Satō (2001) finds that a large majority of individuals in non-regular employment has deliberately chosen this type of employment for reasons such as more flexible working hours. Using data from a 1996 government survey he finds that only 25% of young part-timers and dispatch personnel would have preferred another work arrangement; the figure is as low as 10% for married women and elderly part- timers.
2 http://www.oecd.org/els/soc/LMF1.5%20Gender%20pay%20gaps%20for%20full%20time%20workers%20-
%20updated%20290712.pdf (15.06.2014).
Thus, an almost perfect consensus of academic and public opinion points to an overall negative development of the Japanese labour market over the past two decades. However, there are two general observations that contradict the idea of “two decades lost in the labour market” from the perspective of economic theory.
Firstly, unemployment in Japan has remained very low as compared to most other advanced economies.3 This suggests that labour demand was hardly ever short of labour supply; on the contrary, a number of sectors should have seen supply falling short of demand, namely in small- and medium-sized manufacturing enterprises. It seems, therefore, difficult to imagine how Japanese employers could have realized a “de- standardization” on an aggregate scale. Secondly, where gender wage gaps allow for purchasing the same level of productivity at lower wages, female employment should have increased strongly, possibly even at the expense of men.
In our view, three questions are key to a complete assessment of labour market performance in Japan. Firstly, taking men and women together, is there proof for the claim that regular employment was replaced by non- regular employment on an aggregate level in general, and at the expense of younger generations in particular (Imai 2004)? Secondly, how do today’s labour market outcomes compare to the heydays of the Japanese economy; i.e., to the late 1980s in terms of absolute numbers of employees? Thirdly, have wage gaps in the first two dimensions (gender and standard vs. non-standard employment) actually increased, stayed the same, or have they even shrunk? This last question is key to determining whether the increase in non-regular employment was predominantly due to a shift in labour supply, as is frequently alleged, or rather due to a shift in labour demand.
For our analyses we have set the investigation period to 1988 to 2010. The time frame chosen comes with two specific advantages. First, covering a time span of more than two decades allows for identifying secular trends. Second, relying on 1988 and 2010 as start- and endpoints also helps to ensure that no outliers distort the overall picture in this longitudinal analysis. For instance, an analysis referencing 1997 as the end-point of an exceptional economic phase characterized by over-hiring4, namely the bubble economy, would not only be too short for making inferences on long-term developments, but would also strongly distort results. For the convenience of readers interested in specific periods within the investigation period, the authors have tried to provide annual figures wherever possible.
In our search for answers to our three questions, we proceed as follows. In section 2, we discuss the results of previous studies where relevant insights are provided. Next, section 3 draws on government data to look at how the absolute numbers for employment have developed over the past two decades (section 3.1). Taking into consideration changes in the size of the active population aged 15-64, we then compare 2010 labour market outcomes with a 1988 projection that keeps our two dimensions of gender and standard vs. non- standard employment constant (section 3.2). In both sections, we rely on separate analyses along the two dimensions, as this allows for determining whether regular employment in Japan was substituted with non- regular employment, or whether non-regular employment simply outgrew regular employment. As a third
3 http://www.stat.go.jp/english/data/roudou/results/month/index.htm (15.06.2014)
4
dimension, we then propose an analysis of intergenerational effects in order to gather evidence for the alleged particular hardships among Japanese youth (Ishiguro 2008) and among younger men (Brinton 2011).
Finally, section 4 attempts to shed light on real wage development in order to determine how wage gaps have changed during the investigation period; and thus to understand the predominant cause for the surge in non- regular employment.
2. Previous studies and research objective
We divide previous studies on the shift to non-regular employment into two groups: first, studies that inquire into the nature of the relationship between regular and non-regular employment, and second, studies that analyse the reasons behind the shift to non-regular employment.
To the authors’ best knowledge, there is only one study that has tried to capture overall labour market development by focusing on the question of whether the long-term recession since the 1990s resulted in
“good jobs” (characterized by high wages and benefits, job security, opportunity for training and development) being replaced by “bad jobs” (characterized by the lack of these attributes; Katō and
Kambayashi 2013). Contrary to popular belief, they show that there has not been a secular trend toward non- standard male employment in Japan. Overall, female employment has undergone a compositional change from self-employment to non-standard employment (i.e., from “bad jobs” to “bad jobs” rather than a shift from “good jobs” to “bad jobs”). In the case of young Japanese workers, especially young women, however, they find evidence for a shift of “good jobs” to “bad jobs”. Yet, as the authors rely on relative measures in presenting their findings, they do not provide a basis for answers on whether standard employment has been replaced by non-standard employment in an aggregate perspective.
A number of studies have approached the question of substitution vs. employment expansion from an industry perspective. For instance, Ono (2001) analyses how the use of non-regular employment changed over the years 1989-1999 at two operators of large retail chains. She finds that the two chains increasingly made use of non-regular employees. The share of non-regular employees at chain 1 increased from about 65% to 75%, while that at chain 2 rose from 55% to about 77%. At both chains this increase came at the expense of regular employees. For the same industry, Miyamoto and Nakata (2002) analyse panel data of eleven large department store operators and ten large supermarket operators with more than 1'000 employees between 1987 and 1999. While they also find that standard employment was partially replaced by non- standard employment, they were able to show that an increasing ratio of non-standard employment actually mitigated further downsizing of standard employment.
Yamaguchi (2011) uses microdata from companies listed on Section 1 of the Tokyo Stock Exchange for the years 2005 and 2006. He concludes that regular and non-regular employment in his sample of large and very large companies entertain a substitutive relationship. However, compared to 2005, the elasticity of
substitution decreases in 2006, indicating that fewer regular employees were replaced by non-regular employees. In 2005, a substitutive relationship can be found for the manufacturing, retail, service, construction, and transportation industries, while in 2006, it was limited to the retail, service, and
construction industries. While the study offers interesting insights, it does not lend itself to explaining either
more long-term or more general developments in the Japanese labour market, due to its sample being subject to a twofold selection effect5.
Using a representative sample of employer level microdata from the Koyō Dōkō Chōsa (Survey on Employment Trends; Ministry of Health, Labour, and Welfare), Ishihara and Genda (2003) focus on the relationship between part-time and full-time employment on the micro level between 1991 and 2000. They find that only 3% to 6% of all companies increased their use of part-time employees while reducing regular employment. They further report that only about half of all companies make use of part-time employees, and that companies with fewer than 30 employees tend to employ fewer part-time employees than larger ones.
They also find that throughout the 1990s, the largest losses of full-time employment were seen at companies that made no use of part-time employees at all. What is more, full-time employment was increasingly lost at companies in which part-time employment was also reduced. Relying on data from the same survey, Ishihara (2004) investigates the relationship between full-time and part-time employment between 1991 and 2000.
She finds that an increase in part-time employment did not happen at the expense of full-time employment in over 80% of the cases where full-time employment was lost, and that in over 50% of the cases where part- time employment was increased, it was done regardless of whether full-time employment was reduced or not. Finally, Hara (2003) finds similar results using company-level data from the 1998 Kigyō No Fukuri Kōsei Seido Ni Kansuru Chōsa (Survey of Company Social Welfare Programs; Japan Institute of Life Insurance). While her study covers only the first of the two “lost decades”, her findings clearly point to a complementary relationship.
To sum up, existing research has found mixed evidence on the question of whether standard employment was replaced by non-standard employment or not. Yet, most of the more representative studies using microdata (Miyamoto and Nakata 2002; Ishihara and Genda 2003; Hara 2003) point to a complementary relationship. Contradictory evidence in the studies of Ono and Yamaguchi (2001; 2011) was based on weak methodology. However, neither of these works allows for directly drawing conclusions regarding the extent and segments of the labour market that have seen employment expansion over the last two decades. To shed full light on the topic, an analysis of aggregate employment data for the entire period spanning the “two lost decades” (section 3) is indispensable.
But what are the reasons for the relative shift to non-regular employment? At its heart, the quest for the causes of this development is reduced to the interplay of labour demand (by employers) and labour supply (by employees) in a dual labour market (standard vs. non-standard employment). While inconclusive regarding the primary causes (changes in demand or in supply), the existing literature abounds with hypothetical reasons. Demand side factors such as the intensified competition Japanese companies find themselves entrenched in, slower economic growth (Farber 2007), changes in industrial structure toward a higher share of the tertiary sector (Ishiguro 2008), the adoption of neo-liberal policies, the increased spread of information technology (Suzuki et al. 2010), changes in institutional setting and the behaviour of companies, i.e., the collapse of the “Japanese employment system” (Suzuki et al. 2010) can be found as
purported reasons, just as supply side factors like a growing necessity to supplement household income through female part-time employment (Kawata and Naganuma, 2010), government policy encouraging women to work on a part-time basis and people's diverse working preferences (Ishiguro 2008; Yamada 2011), as well as increased individualization of career design and life plan (Suzuki et. al. 2010) have been suggested.
Interestingly, changing patterns of the latter (labour supply) are frequently considered more important factors than change of the former. Unfortunately, the literature falls short of providing a clear-cut link to wage developments in the respective segments. However, a first clue to this inter-relationship can be found in a study of Hayami and Matsuura (2001), which computed the elasticity of substitution between full-time and part-time workers using macro time series from 1982 to 1997. They find elasticity of substitution of 5.32 for the manufacturing industry, and 5.87 for the non-manufacturing industries. This means that a 1% change in relative wages (i.e., the wage gap) results in a 5 to 6% change in the relative employment ratio between full- time and part-time employment, implying that a small change in relative wages leads to large substitution effects. From this we can note that an increase in the wage gap between regular and non-regular employment suggests an increase in the rate of substitution between the two types of employment. Consequently, an analysis of the development of wage gaps during the “lost decades” (section 4) is needed to understand whether supply side factors (widening gap), or demand side factors (closing gap) were the primary cause for the relative shift toward non-standard employment.
3. Structural change in the Japanese labour market
We start in section 3.1 by analysing labour force participation rates and the relative share and absolute numbers of standard and non-standard employment, overall and by sex, in order to obtain an understanding of the magnitude of changes in the structural parameters of the Japanese labour market. Findings obtained from this analysis will help to evaluate the absorptive capacity of the labour market in section 3.2.
Subsequently, in section 3.3 we aim at evidencing changes for specific groups that are frequently referred to both in the literature, and in public discourse. To that end, we provide a detailed analysis by age-cohorts and by sex.
Basing our study on macro data obtained from the Labour Force Survey will likely provoke two specific objections. First, the choice of macro data might not seem entirely appropriate. Yet, this methodological choice was guided by the observation that the on-going discussion seems to lack a solid base that would allow for putting into perspective the many individual findings obtained from studies on particular aspects of the labour market such as the gender gap, or developments in certain industries.
Second, critics might object that government statistics do not appropriately reflect the 1998 change in labour laws that has led to counting employees with fixed-term contracts of more than one year as regular
employees. The share of individuals concerned, however, is minuscule: the proportion of standard employees (both sexes) with fixed-term contracts as a percentage of the population aged 18-70 declined from 0.5% in 1982 to 0.2% in 2007. When analysed by sex for the same time frame, the proportion of male employees was reduced from 0.6% to 0.3%, and that of female employees from 0.3% to 0.1% (Katō and Kambayashi 2013).
3.1 A matter of perspective: relative and absolute measures of regular and non-regular employment Between 1988 and 2010 the share of regular employees has declined from 83.3% to 69.9% for the active population aged 15-64, while the corresponding share of non-standard employment has increased from 16.7% to 30.1%. When analysed by sex, the share of men in non-standard employment has increased from 6.8% in 1988 to 14.4%, while that of women has risen from 33.9% to 50.9%. At first sight, these figures appear to match with the negative public perception and with most existing research on the structural development of the Japanese labour market.
An analysis of absolute employment numbers for individuals aged 15-64, however, yields a more
differentiated picture. Between 1988 and 2010, the total number of employees increased from 43.2 million to 51.7 million (+8.5 million / +19.7%). A very large share of this increase consists of non-standard employees:
the number of non-standard employees rose by 8.4 million, from 7.2 million to 15.6 million, which
corresponds to an increase of 115.8%. While this seems to confirm the trend towards a de-standardisation of work in the Japanese labour market, the fact that the number of regular employees also slightly increased in the same time frame, from 36.0 million to 36.1 million (+0.1 million / +0.4%), is frequently ignored and leads to a first important implication: on an aggregated basis, no standard employment has been replaced with non-standard employment (compare Table 1).
Conducting the same analysis separately by sex yields further insightful results. The total number of male employees aged 15-64 increased from 27.4 million to 29.4 million (+2.0 million / +7.6%). With non-standard positions accounting for more than the total of this increase (+2.4 million / +127.3%), this leaves a negative margin to the number of regular male employees (-0.3 million / -1.2%). While the increase in the number of female employees from 15.8 million to 22.2 million (+6.4 million / +40.6%) shows a similar pattern, with the number of non-regular employees rising from 5.4 million to 11.3 million (+5.9 million / +111.2%), the increase in the number of regular female employees from 10.5 million to 10.9 million (+0.4 million / +4.4%) occurred with a considerable positive margin. Putting this evidence together, we can argue that not only were all regular jobs newly created between 1988 and 2010 obtained by women, but also, that women have succeeded in making inroads into male standard employment. In light of the critical discussion on the role and position of female employees in the Japanese labour market, this is a rather surprising result. Compare Table 1 for the corresponding data.
Table 1: Regular and non-regular employees aged 15-64 (in millions), 1988-2010, total and by sex
Source: Based on MHLW (2012a); Note: Table may contain rounding errors.6
Naturally, an analysis of absolute employment numbers cannot go without taking account of changes in the overall population. Between 1988 and 2010, the Japanese population aged 15-64 decreased from 85.0 to 81.7 million (-3.9%). If we now combine these figures with the above findings on absolute employment numbers, we can see that the decrease in the population aged 15-64 of -3.9% stands in stark contrast to the increase in the number of employees (+19.7%). This translates into a strong increase in the country's employment ratio (+12.4pp). The same also holds true when analysed separately by sex: The number of male employees increased by 7.6% despite a 3.3% decrease in the male population aged 15-64, translating into a 7.2pp increase in the male employment ratio. The number of female employees increased by as much as 40.6%, despite a 4.4% decrease in the female population aged 15-64; hence, the female employment ratio increased by a hefty 17.5pp (compare Table 2).
Table 2: Population and employed persons aged 15-64 (millions), employment ratios (%)
Source: Own calculations based on data from the Statistics Bureau Japan, Historical Statistics, Chapter 2 (Population and Households), Table 2-3, and on MHLW (2012a);
Note: Table may contain rounding errors.
6 Rounding errors occur when rounded-off values are used instead of exact mathematical values. For instance, in Table 2, 2010 rounded values for men in regular employment (25.2 million) and non-regular employment (4.3 million) would add up to a larger total (29.5 million) than the equally rounded value displayed (29.4 million).
To summarize, the Japanese labour market has accommodated 8.5 million new employees aged 15-64 by creating 8.4 million non-regular jobs and 0.1 million regular jobs between 1988 and 2010. On an aggregated basis, no standard employment was replaced with non-standard employment. Of the 8.5 million newly created jobs, 6.4 million went to women. Although only 0.4 million out of the 6.4 million new jobs for women are in standard employment, these 0.4 million equal 400% of the net increase in standard
employment of 0.1 million, i.e., women have secured more regular employment at the expense of men. The question of which group(s) of women gained regular employment at the expense of which group(s) of men will be dealt with in section 3.2.2.
3.2 Job creation capability of the Japanese labour market 1988 to 2010
We further assess the performance of the Japanese labour market in terms of job creation capability by calculating how many employees the Japanese labour market would have had to accommodate by 2010 if
• Employment ratios (number of employees relative to active population), and
• Shares of standard vs. non-standard employment
had remained constant at the 1988 level for men and women aged 15-64. Comparing the results of this model with the actual 2010 figures will allow us to assess the absorptive capacity of the Japanese labour market during the investigation period.
Labelled as “Model 1988”, Table 3 lists the numbers of regular and non-regular jobs for men and women that were to be expected if both employment ratios and the ratios of standard and non-standard employment had stayed constant between 1988 and 2010. Calculating the differences to the actual figures as shown above in Table 4, we understand that 10.1 million more jobs have been created than were to be expected from a 1988 perspective, given the labour market structure at that time.
Table 3: “Model 1988” versus 2010 actual figures (millions)
Source: Own calculations based on data from the Statistics Bureau Japan, Historical Statistics, Chapter 2 (Population and Households), Table 2-3, and on MHLW (2012a); Note: Population aged 15-64; table may contain rounding errors.
What is more, and most surprisingly, 1.4 million of the jobs created beyond the 1988 model are jobs in standard full-time employment. When analysed by sex, the Japanese labour market created 0.5 million more regular jobs for men, and 0.9 million more regular jobs for women, confirming the increasing inroads of women into the Japanese labour market.
3.3 Intergenerational inequality? – Standard employment by age cohorts, 1988 to 2010
The results obtained in sections 3.1 and 3.2, however, do not provide evidence on the extent of inter- generational effects, namely on the reported difficulties of middle/high school graduates and university graduates to find permanent employment, nor do they reveal the characteristics of those women who have made inroads into the Japanese labour market. Therefore, we now propose to analyse the development of the share of standard employment for different age cohorts during the time period 1988 to 2009. Note that we do not analyse the share of standard employment relative to the labour force for two reasons. First, the labour force is sensitive to business cycles, while age cohorts are not. Second, the definition and thus the size of the labour force can change, complicating the analysis of long-term trends or rendering them altogether
impossible to identify.
As can be inferred from Figure 1 and Annex 1, a substantial decline in the share of standard employment can only be observed within the age cohorts of (a) 15 to 24 years for both sexes, and of (b) 25 to 34 years for men.
Figure 1: Share of standard employment by age cohorts, 1988-2009
Source: Own calculations based on data from the Statistics Bureau Japan, Historical Statistics, Chapter 2 (Population and Households), Table 2-3, and MHLW (2012a).
The finding of a decreasing share in standard employment, noted as (a) and (b) above, seems to corroborate the claims of a “lost generation” - and hence, of poor labour market performance. As they obviously mark exceptions to a general trend, we will enquire into the two cases in more detail.
3.3.1 Disadvantaged Japanese youth?
According to Suzuki et al. (2010), the employment gap among young workers started to expand around the year 2000. Companies cut back on hiring new regular recruits, but instead hired more non-regular workers.
regular employment with little opportunity to grow, either in terms of skills or wages.
The claim of an increasingly disadvantaged youth, however, is not supported when employment numbers of the age cohort 15 to 24 are analysed: the absolute number of 15-24 year-old non-regular employees
decreased from 2.59 million in 2002 to 2.23 million in 2010 (-13.9%) (compare Table 5). When expressed as a percentage of the corresponding age-cohort, in 2002 17.0% of the 15-24 year-olds were employed as non- regular employees, and 17.7% in 2010 (compare Table 4). Since 2002, the data on non-regular employees distinguishes two groups: non-regular employees including persons attending school, to whom the above numbers refer, and non-regular employees excluding persons attending school. The numbers reveal that in 2002, 45% of the non-regular employees in this age bracket were students in tertiary education, i.e., working part-time while attending school. This figure rose to 50% in 2010. Now, this means that the absolute number of non-regular employees between 15 and 24 years who are not concurrently attending school decreased from 9.3% in 2002 to 8.9% in 2010 (Table 4). Before this background we understand that two factors can be held responsible for the decrease in the number and share of regular employees within this age cohort (Figure 1). First, the decrease in the share of regular employees within the respective age cohort of 15 to 24 can be accounted for by a corresponding increase in the average time spent in education. Young Japanese have been continuously enjoying ever higher standards of education over the last 25 years; this can be inferred from the increasing advancement rates between high school and university, which increased by 17pp for both sexes, by 12pp for male students, and by a hefty 23pp for female students between 1985 and 2010 (Statistics Bureau of Japan, Historical Statistics 25-12 “Enrolment Advancement Rates (1948-2005)”
http://www.stat.go.jp/english/data/chouki/25.htm). Second, while the overall development was positive, certain low-skilled youths (middle/high-school graduates) have increasingly faced difficulties in their transition from school to work. Genda et al. (2010) found that the worse labour market conditions were at entry, the lower was the ratio of regular employees among less well-educated Japanese men, i.e., male high- school graduates. This finding, i.e., that difficulties in the transition between high-school and workplace have increased for young Japanese males, accords with the insights of Brinton (2011), who conducted in-depth interviews with a sample of male high-school students who had finished school in the depth of the Japanese recession and faced considerable difficulties in the transition from school to work. The same applies to young women with a junior and senior high-school education, for whom it has also become less likely to find standard employment, as demonstrated by Abe (2010). With this, we understand how increasing educational levels – on the one hand – allow for explaining a seeming increase of non-regular employment, and – on the other hand – create increasing competition on the labour market for low-skilled entrants.
Table 4: Share of age cohort 15–24 employed as non-regular employees (age cohorts in 10'000 persons)
Source: Own calculations based on Statistics Bureau of Japan, Labour Force Survey, long-time series, Table 10 (http://www.stat.go.jp/data/roudou/longtime/03roudou.htm) and Statistics Bureau of Japan, Population Estimates, long- time series (http://www.e-stat.go.jp/SG1/estat/List.do?bid=000001039703&cycode=0)
Table 5: Absolute numbers of non-regular employees (2002-2012, ) in the age cohort 15-24 (in 10'000 persons; annual averages)
Source: Own calculations based on Statistics Bureau of Japan, Labour Force Survey, long-time series, Table 10
Data on the development of the number of vacancies for university graduates in private industry (Works Institute 2012) sheds additional (normatively positive) light on the job situation for the age bracket of 15-24.
The number of open (regular) positions for university graduates in private industry (all company sizes) increased from 608'000 in 1987 to 725'300 in crisis-struck 2010, and, except for the year 2000, it has always exceeded the number of applicants (259'500 in 1987, 447'000 in 2010). The number of job openings in large firms, which have traditionally been more sought after by university graduates, more than doubled between 1996 and 2010, from 64'500 to 159'700. Rather steep increases took place in 1997 and 1998, when the Japanese economy regained some momentum, and during the economic recovery that started in 2003 and came to a halt in 2009.
This halt was the result of the U.S. real estate crisis following the 2008 Lehman Shock.7 The ensuing global economic downturn had detrimental effects on those companies in Japan that are typically ranked among the most favoured employers by university graduates: export-driven large companies. As the decline in demand from overseas markets was eventually compounded by a strong appreciation of the Yen from 2010 to 2012 (compare Blind and Lottanti von Mandach 2012), many of these favourite employers sharply curbed or entirely stopped hiring new graduates, even revoking contracts already issued to the class of 2009.
Occurrences such as the previously unknown revocation of contracts caused widespread insecurity among the general public, which in recent years has effectively precluded an objective perception of the labour market by young graduates.
3.3.2 Males aged 25 to 34 facing stronger competition for standard employment – by women
As we have just seen, job openings are still outnumbering new graduates by a significant multiple in spite of the strong expansion of tertiary education over the last two decades. With plenty of standard employment available for university graduates despite the 2008 Lehman Shock (Works Institute, 2012), i.e., for well over 40% of every cohort since 2000, and over 50% since 2010 (MEXT, 2013), the outlook should not be too grim for those advancing into the 25-34 age bracket. As can be seen in Annex 1, total standard employment relative to this particular age cohort has almost remained constant with a minor decrease of only 0.6pp (from 56.7% in 1988 to 56.1% in 2009).
We find that male standard employment in the age group of 25-34 decreased significantly by 9.0pp, from 82.2% in 1988 to 73.2% in 2009. However, with the total decrease at less than 1pp, it was a marked increase in female standard employment (by some 8.1pp from 30.3% to 38.4%) that almost entirely compensated for the decline among their male peers. In other words, younger women are increasingly finding their way into standard employment at the expense of young men (compare Annex 1).
If noted at all, the appraisal of this development has been far from positive. First, the advancement of female university graduates at the expense of their male peers has fuelled fears that it will negatively affect the
7 Financial services firm Lehman Brothers filed for Chapter 11 bankruptcy protection on September 15, 2008.
Lehman's bankruptcy filing is regarded to have played a major role in the unfolding of the 2008 global financial crisis.
(chronically low) fertility rate in Japan for two reasons: (1) because women are forced to choose either career or family (which Arai and Lechevalier 2005 call the “polarization of choices” for women) and (2), because Japanese men in non-standard employment have (more) difficulties finding a wife and starting a family compared to their peers in standard employment, because they cannot live up to the ideal of the male breadwinner (Mikanagi 2011). While the latter indeed seems to be the case (Miura 2007), with regard to the first reason, and contrary to expectations, female regular employment actually seems to promote rather than discourage marriage, as Nagase (2002) has shown.
Second, female labour market advancement has accentuated inequalities among women, as Tachibanaki (2008) points out. Whether women are hired as regular or non-regular employees, or, in the case of regular employees, whether they are on the management/career track (sōgō shoku) or clerical/non-career track (ippan shoku) has a significant effect on working conditions, possibilities for promotion, awarding of employee benefits, participation in a social insurance program, and, last but not least, wages. As a matter of fact, the rate of female managers at section head manager-level more than quadrupled between 1982 and 2010, but is still in the single-digit percentage area (1982: 1.4%; 2010: 6.2%; compare http://www.works-
i.com/pdf/r_000327-4.pdf). Obviously, much remains to be done.
4. Real wage development 1988-2010: Gaps closing in two dimensions
As we have shown in section 3, recent labour market expansion has been dominated by an increase in non- regular employment, especially in female non-regular employment. This increase is a result of significantly higher labour market participation rates, especially by women. The extant literature frequently attributes this development to factors resulting in a shift in labour supply: first, to the government's ongoing policy of en- couraging women to work on a part-time basis (Ishiguro 2008; Yamada 2011), and second, to a growing necessity to supplement household income through female part-time employment (Kawata and Naganuma 2010).
However, earlier evidence for a significant impact of government policies on female labour market
participation rates, both in regular and in non-regular employment, e.g., Yamada (2011), should be regarded with some skepticism in the light of more recent research (Steinberg 2012). The second explanation refers to the so-called “household assistance effect” (also known as the “Douglas-Arisawa Law”), stating that the probability of a wife's labour market participation is negatively correlated with the husband's income level. A number of theoretical and empirical analyses demonstrated the validity of this law in the 1960s and 1970s.
Later studies on Japan, however, produced mixed results. A most recent panel data analysis of changes in female labour supply in Japan for the period 1993-2008 concludes that the household assistance effect still holds true, even for younger aged cohorts (Kishi 2012). While this might partly explain the increasing numbers of female non-regular employees (Kawata and Naganuma, 2010), it does not, however, explain the increase in male non-regular employment. We reckon that the observed outcome might instead be due to a shift in labour demand for non-regular employees, with rising wages in non-standard employment attracting
an increasing number of employees, both male and female. Correspondingly, we provide an in-depth analysis of real wages by age-cohorts and by sex in order to assess how wage gaps between men and women, and between standard and non-standard employment, have developed in the course of the last two decades.
For a start, however, we first look into the development of the aggregate wage share. As it turns out, indivi- duals in dependent employment have succeeded in securing an increasing share of net national income through their wages from about 60% in 1988 to 66% in 2010 (CAO 2012). While this increasing share is ar- guably due to lower returns on capital investments rather than to successful union bargaining, one can infer from this that Japanese employers have by no means robbed dependently employed individuals of the fruits of their work.
Critics may argue that an ever increasing share of the population is required for reaping the fruit: as we have seen, the employment ratio of individuals aged 15 – 64 has increased by 12.4pp from 50.8% in 1988 to 63.2% in 2010 (Table 2). Equalling a relative increase of 24%, such argument deserves closer inspection.
The Japanese Employment Status Survey also provides the sum of total hours worked per year (excluding agriculture): between 1988 and 2010 we can observe a 8.3% decline in total hours worked, substantially more than the 3.9% decrease in the active population aged 15-64. As the decrease in employment in the agricultural sector was certainly even more pronounced, we can safely argue that the fruit can be reaped with less effort nowadays.
We now analyse wage developments of male and female employment, as well as in standard and non- standard employment. While there are naturally many empirical factors that can impact on real wages, such as the deregulation of industries (compare Kawaguchi and Mizuno 2011), we here sketch out the theoretical effects of demand and supply shifts to be expected in a dual labour market, because, as we mentioned in our introduction, at the very heart of the relative shift to non-standard employment is the interplay between labour demand (employers) and labour supply (employees) with a clear-cut link to wages. Figure 2a illustrates the wage gap Δw0 in a dual labour market.
Figure 2a: Wage gap in a dual labour market
In the case of a demand shift away from regular and towards non-regular employment, microeconomic reasoning suggests that prices for substitutes will converge toward the cheaper alternative. Below, Figure 2b applies this reasoning given non-elastic labour supply in the two segments. As becomes evident from this graphic analysis, the initial wage gap Δw0 is expected to decrease; i.e., Δw1 < w0.
Figure 2b: Convergence of wages due to demand shift in the labour market
Next, we sketch the effect on wages to be expected from a shift in labour supply, i.e., the effect on wages that should be expected from a sudden increase in labour supply (e.g., through the need to supplement household income). Figure 2c applies this reasoning given non-elastic labour demand. As becomes evident from the graphic analysis, the initial wage gap Δw0 is expected to increase in this case; i.e., Δw1 > w0.
Figure 2c: Divergence of wages due to supply shock in the labour market
In order to test the effects illustrated in Table 2b and c, we analyse the data on real wages earned by men and women in standard and non-standard employment between t1 = 1988 and t2 = 2010. We use data from government statistics obtained through the Basic Survey on Wage Structure (MHLW 2012b). Wages for regular employees include bonuses, but exclude overtime payment; for non-regular employees we have used hourly wages. We index8 all wages to 1988 for the 2x2 matrix that results from separating men from women and standard from non-standard employment and plot the results in Figure 3.
Figure 3: Wage development standard vs. non-standard employment, male vs. female, age cohorts 20-64 (1988-2010, 1988=100)
Source: Own calculations based on MHLW (2012b).
Two observations are most striking: First, increases in real wages of non-regular employees were equal to or larger than those of regular employees for both men and women. We can thus confirm the theoretical
prediction sketched in Figure 2b for a demand shift from regular to non-regular employment. It also becomes obvious that this shift in demand effectively was stronger than the supply effect implied by the tax incentives that Japanese lawmakers tied to part-time employment, and thus to female non-regular employment. Second, the narrowing of wage gaps was particularly pronounced in the case of female non-standard employment, where certain age groups saw very strong increases (all groups between 30 and 49 at +40% or more) (Figure 4).
8 In comparing the development of prices (wages as prices of labour) over time, indexing to a specific base year helps to identify diverging developments. Commonly, base values are set to 100 as changes during the observation period can be read as percentages.
Figure 4: Wage development: standard vs. non-standard employment, male vs. female, by age cohorts (1988-2010, 1988=100)
Source: Own calculations based on MHLW (2012b).
With this as background, we can discern that two frequently deplored wage gaps have been narrowing significantly in Japan since the late 1980s. First, the male vs. female divide: Young women have made further inroads into the Japanese labour market and have not only increased their share in terms of standard employment, but also in terms of real wages. And second, the standard vs. non-standard divide: While a significant wage gap continues to exist between regular and non-regular employees, this gap has sharply narrowed. Our findings also bound well with the strongly positive wage elasticity of substitution between regular and non-regular employment estimated by Hayami and Matsuura (2001) for the time span between 1982 and 1997. In light of the significant employment expansion that we evidenced in section 3, we
understand that much of the rise in non-regular employment can be explained as the positive response of the labour force to increased demand for and to corresponding wage increases in non-standard employment.
Enlarging analytical scope to also include the intergenerational dimension, we find no significant evidence for a trend toward more intergenerational inequality (Figure 4). What is more, our findings on the first two dimensions extend to virtually all age cohorts. Closing wage gaps between regular and non-regular employ- ment can be found in all age groups with the sole exception of men between 30 and 34; also, increases in female wages have been equal to or markedly larger than those of their male peers across all age cohorts.
5. Concluding remarks
We have assessed the state of the Japanese economy between 1988 and 2010 by taking a fresh analytical look at its labour market in an aggregate perspective. Our study was motivated by two observations. First, the ongoing discussion both in academia and public forums seems to have lacked a solid basis encompassing the many individual findings obtained from qualitative and quantitative studies on particular aspects of the labour market. Second, no study has provided answers to three questions that are key to an overall
assessment of labour market performance. The first of these questions was whether regular employment was replaced by non-regular employment on an aggregate level in general, and at the expense of younger
generations in particular. Second, we wanted to know how today’s labour market outcomes compare to the heydays of the Japanese economy, i.e., to the late 1980s, in terms of absolute numbers of employees. Finally, it was imperative to know whether wage gaps between men and women as well as regular and non-regular employees have increased, stayed the same, or shrunk. This last question was key to determining whether the increase in non-regular employment was predominantly due to a shift in labour supply (employees) or due to a shift in labour demand (employers).
Our findings strongly indicate that the label “lost decade(s)” cannot be corroborated. From our quantitative analyses we find no evidence of any hollowing out of the Japanese employment system. Contrary to both public perception and many research opinions, we find that the Japanese economy as a whole has seen highly stable employment conditions in a long-term perspective. Our main results can be summarized as follows.
First, between 1988 and 2010, a total of 8.5 million new jobs were created for those aged 15-64, including 0.1 million in regular employment. Thus, on an aggregate basis, standard employment was not subject to replacement by non-standard employment. Naturally, this finding does not deny individual cases
documented for specific industries, or, the existence of individual firms where regular employment was effectively replaced by non-regular employment (as evidenced, e.g., in Ono 2001, Miyamoto and Nakata 2002).
Second, we find a strong increase in employment ratios of the active population for both sexes during the same period. Thus, the Japanese labour market has accommodated an increasing share of the population in an economy that has been growing only slowly since 1991. In order to assess the Japanese labour market in terms of absorptive capacity, we calculated how many jobs would have had to be created if the 1988 employment ratios of both sexes had stayed constant, and compared this number to the actual number of newly created jobs. We find that between 1988 and 2010, Japanese businesses created 10.1 million more jobs than were to be expected given the 1988 structural parameters of the labour market (participation rates and shares of regular vs. non-regular employment); what is more, 1.4 million of this unexpected increase were jobs in regular employment.
Third, we find that the share of regular employees relative to the active population aged 15-64 remained constant. This also holds for an analysis by age cohorts, but with two exceptions: (i) In the age cohort 15-24 for both sexes, we find a marked increase in the share of non-standard employment. This increase, however, can almost entirely be explained by increasing time spent in education associated with part-time jobs. (ii) For men in the age cohort 25-34, we confirm the frequently reported decrease in the share of regular jobs (e.g., Cook 2013). However, this decline of about 10pp over the last 20 years is almost entirely compensated for by a corresponding increase in female regular employment. Thus, in our view, this finding suggests that Japanese young women have, at least from a quantitative perspective, made substantial inroads into the Japanese labour market.
Fourth, and most significantly, women have succeeded in securing 400,000 additional jobs in standard employment. With the net increase at only 100,000 this “female success” came at the expense of male standard employment. In spite of this promising development, these quantitative successes do not make up for the many remaining gender issues of a more qualitative nature in the Japanese workplace (compare Kim 2008).
Fifth, our analysis of real wage development further confirms female labour market achievement.
During the observation period female real wages have seen stronger increases than those of men across all age-cohorts (20-64) for non-standard employment and for all cohorts up to 55 in standard employment (wage gains were about equal for men and women aged 55+ in standard employment).
Sixth, our wage analysis further provides evidence that the wage gap between non-standard and standard employment narrowed significantly during the observation period, thus confirming the predictions of micro-economic theory as to the effects of a demand shift from standard to non-standard employment. In other words, the increase in non-regular employment was predominantly due to an increase in labour demand by employers.
In light of these normatively positive findings, the question arises why popular and academic discourse overwhelmingly share a negative perspective on labour market developments. While such inquiry certainly
explanation here.
First, it could be argued that a worsening of the Japanese labour market resonates well with the new societal model of Japan as a divided society (kakusa shakai; Tachibanaki 1998) that has come to replace the
previously prevailing general middle-class society model (sōchūryū shakai). In order for a societal model to be dominant, however, it does not necessarily have to reflect the actual social structures. In fact, empirical studies fail to support either of these two societal models (Chiavacci 2008). Rather, in order to attain and retain a dominant position, a model needs to connect to the mainstream perception. It then functions as a frame through which everyday experiences are assessed and has a crucial impact on social actors by shaping their perspective. It is important to note that the establishment of particular societal models is also contingent upon ongoing political struggle. The dominance of a new societal model of inequality may thus explain why both popular and academic discourse tend to largely ignore any counter-evidence, such as absolute
employment numbers, or the increase in female and student labour market participation, which can mainly be attributed to a labour market pull, as we have shown.
A similar bias can be observed with regards to household incomes. Both popular and academic discourse reiterate the notion that inequality between household incomes has been widening in Japan and take this as (further) proof that the Japanese society has turned into a “gap society” (kakusa shakai). Most accounts, however, make no reference to the fact that the observed increase in income inequality can actually be attributed to two factors: population aging and changes in household structures (compare Ōtake 2005).
An additional line of explanation may be derived from what macro economists refer to as ‘adaptive expectations’ (Owen and Honkapohja 2001). In contrast to the received model of rational expectations that are seen always to reflect the current state of an economic system, the formation of ‘adaptive expectations’
imply a significant time lag. As Kamada and Muto remark, concepts of rational expectations “fail to deal with changes in long-run equilibrium” (2000: 23). This would explain why evaluations of labour market performance still take all-time highs such as 1997 as their reference9.
As a side note, in the case of female labour market advancement or, better, its lack of appraisal, the
conservative establishment has actually sided with its political enemy. The ongoing strong prevalence of the societal ideal of a male breadwinner in certain layers of Japanese society arguably precludes a positive appraisal of young women increasingly finding their way into standard employment, especially where this happens at the expense of men. This comes in spite of the fact that double income couples outnumbered households with the “male breadwinner”-model as far back as the 1980s (Shire 2008). Recently, current prime minister Abe Shinzō has made a plea to increase the labour market participation of women as part of his Abenomics. While arguably not more than a “second-worst” way of avoiding large-scale labour immigration, it will be interesting to observe whether Abe's attempts to put “woman power” to work and make Japan “a place where women shine” will eventually result in a positive appraisal of female labour market advancement.
9 Although land and stock prices had already started to fall in 1990, firms, trapped in their planning routines and anticipating that high growth rates would eventually return, continued to hire increasing numbers of regular employees up to 1997 (MHLW 2012a).
What do these findings imply for future labour market prospects in Japan? While still a rather unreliable predictor of the future, the past remains the best available alternative. With this in mind, our findings suggest that Japanese employers will likely continue to respect their social responsibility in keeping implicit lifetime employment promises whenever possible. From that, we may derive (1) that labour productivity gains in Japan will continue to be low and mainly driven by technological improvements and (2) that labour market demand will stay strong. However, with the significant past increases in female labour market participation documented in this paper, maintaining the current pace will require ever stronger efforts. It remains to be seen whether the Abe government will succeed in delivering on its promise.
6. References
Abe Yukiko. 2010. 'Equal Employment Opportunity Law and the Gender Wage Gap in Japan: A Cohort Analysis'. Journal of Asian Economics 21 (2): 142–155.
Antonczyk, Dirk, Bernd Fitzenberger and Katrin Sommerfeld. 2010. 'Rising Wage Inequality, the Decline of Collective Bargaining, and the Gender Wage Gap'. Labour Economics 17: 835–847.
Appelbaum, Eileen. 2002. 'Transformation of Work and Employment and New Insecurities.' In The Future of Work, Employment and Social Protection. The Search for New Securities in a World of Growing
Uncertainties, ed. Peter Auer and Christine Daniel Christine. Geneva: International Labour Organisation (ILO): 17–37.
Arai Misako and Sébastien Lechevalier. 2005. 'The Inequalities between Men and Women in the Japanese Labour Market : A Regulationist Approach'. Keizai kagaku (The Economic Science) 52 (4): 129–154.
http://ir.nul.nagoya-u.ac.jp/jspui/handle/2237/10761 (11.06.2014).
Beck, Ulrich. 1992. Risk Society: Towards a New Modernity. London: Sage.
Bernhardt, Annette and Dave Marcotte. 2000. 'Is „Standard Employment“ Still What It Used to Be?'. In Nonstandard Work: The Nature and Challenges of Changing Employment Arrangements, ed. Françoise Carré, Marianne A. Ferber, Lonnie Golden, and Steve Herzenberg. Champaign, IL: Industrial Relations Research Association: 21-40.
Blind, Georg and Stefania Lottanti von Mandach. 2012. 'Neither Struck by the British, nor the Dutch Diseases: The Japanese Economy Comes Along Quite Healthy'. In Japan 2012. Politik, Gesellschaft, Wirtschaft, ed. David Chiavacci and Iris Wieczorek, Berlin: Vereinigung für sozialwissenschaftliche Japanforschung: 131–149.
Brinton, Mary. 2011. Lost in Transition: Youth, Work, and Instability in Postindustrial Japan. Cambridge:
Cambridge University Press.
CAO (Cabinet Office). 2012. National Accounts, Time Series Data. Tokyo: Cabinet Office, www.esri.cao.go.jp/ (24.11.2013).
Chiavacci, David. 2008. 'From Class Struggle to General Middle-Class Society to Divided Society: Societal Models of Inequality in Postwar Japan'. Social Science Japan Journal 11 (1): 5–27.
Cook, Emma. 2013. ‘Expectations of Failure: Maturity and Masculinity for Freeters in Contemporary Japan’.
Social Science Japan Journal 16 (1): 29–43.
Chusseau, Nathalie and Michel Dumont. 2012. 'Growing Income Equalities in Advanced Countries.' In Growing Income Inequalities. Economic Analyses, ed. Joël Hellier and Nathalie Chusseau, Basingstoke:
Palgrave Macmillan: 13–47.
Diekmann, Andreas and Ben Jann. 2005. 'The Erosion of Regular Work. An Analysis of the Structural Changes in the Swiss and German Labor Markets'. In Contemporary Switzerland: Revisiting the Special Case, ed. Hanspeter Kriesi, Peter Farago, Martin Kohli and Milad Zarin-Nejadan, Houndmills: Palgrave Macmillan: 190–217.
Farber, Henry S. 2007. Labour Market Adjustment to Globalization: Long-Term Employment in the United
Fu, Huiyan. 2011. An Emerging Non-Regular Labour Force in Japan: The Dignity of Dispatched Workers.
London and New York: Routledge.
Genda Yūji, Kondo Ayako and Ōta Sōichi. 2010. 'Long-Term Effects of a Recession at Labor Market Entry in Japan and the United States.' The Journal of Human Resources 45 (1): 157–196.
Gottfried, Heidi and Jacqueline O'Reilly. 2002. 'Reregulating Breadwinner Models in Socially Conservative Welfare Systems: Comparing Germany and Japan.' Social Politics 9 (1): 29–59.
Hara Hiromi. 2003. 'Seiki rōdō to hiseiki rōdō no daitai hokan kankei no keisoku' (Estimating the Elasticities of Complementarity Between Regular and Non-Regular Employees In Japan). Nihon rōdō kenkyū zasshi (The Japanese Journal of Labour Studies) 518: 84–100.
Hayami Hitoshi and Matsuura Toshiyuki. 2001. From 'Life-Time' Employment To Work-Sharing: Fact Findings and an Estimation of Labour Demand for Part-Time Workers in Japan. Keio Economic Oberservatory Discussion Paper No. 65.
Hevenstone, Debra. 2010. 'National Context and Atypical Employment.' International Journal of Sociology 25: 315–347.
Imai Jun. 2004. The Rise of Temporary Employment in Japan: Legislation and Expansion of a Non-Regular Employment Form. Duisburger Arbeitspapiere Ostasienwissenschaften 62/2004.
Ishiguro Kuniko. 2008. 'Japanese Employment in Transformation. The Growing Number of Non-Regular Workers.' Electronic Journal of Contemporary Japanese Studies, Article 10 in 2008.
Ishihara Mamiko. 2004. 'Pātotaimu koyō no kakudai wa furutaimu no koyō o herashite iru no ka' (Does Part- Time Work Reduce Full-Time Work?). Nihon rōdō kenkyū kōzō (The Monthly Journal of the Japan Institute of Labour) 45 (9): 4–16.
Ishihara Mamiko and Genda Yūji. 2003. 'Jigyōsho reberu de mita pātotaimu koyō no hendō - “Pāto zōka ga sei'shain o herasu” wa hontō na no ka' (Investigating the Change in the Use of Non-regular Employees from a Micro Level – Does the Increased Use of Non-regular Employees Really Reduce Regular Employment?).
Keizai bunseki (Economic Analysis) 168: 83–95.
http://www.esri.go.jp/jp/archive/bun/bun168/bun168.html#toc (15.10.2013)
Itō Shōken. 2012. 'Seichōritsu teika ni tsuite no nōto' (A Note on the Slowdown of Japanese Economic Growth). Gendai shakai kenkyū (Contemporary Society) 15: 99-110.
JILPT (The Japan Institute for labour Policy and Training ). 2011. Non-Regular Employment Issues and Challenges Common to the Major Developed Countries. JILPT REPORT No. 10.
http://www.jil.go.jp/english/reports/jilpt_02.html (3.3.2012)
Kamada, Kōichirō and Mutō, Ichirō. 2000. Forward-Looking Models and Monetary Policy in Japan. Tokyo:
Bank of Japan Working Paper Series: 00-7.
Katō Takao and Kambayashi Ryō. 2013. Good Jobs, Bad Jobs, and the Great Recession: Lessons from Japan's Lost Decade. Institute of Economic Research, Hitotsubashi University, PRIMCED Discussion Paper Series, No. 41. http://www.ier.hit-u.ac.jp/primced/e-index.html (21.10.2013).
Kawaguchi Akira and Mizuno Keizō. 2011. 'Deregulation and Labour Earnings: Three Motor Carrier Industries in Japan.' Labour Economics 18: 441–452.
Kawata Hiroshi and Naganuma Saori. 2010. Labour Force Participation Rate in Japan. Bank of Japan Review 10-E-7, http://www.boj.or.jp/en/research/wps_rev/rev_2010/rev10e07.htm/ (24.11.2013) Keizer, Arjan B. 2008. 'Non-Regular Employment in Japan: Continued and Renewed Dualities.' Work Employment Society 22: 407–425.
Kim, Young. 2008. 'Personnel Management Reforms in Japanese Supermarkets The Positional Warfare and Limited Assimilation of Conversational Communities'. Social Science Japan Journal 11 (2): 183–199.
Kishi Tomoko. 2012. Changes in the Labour Supply of Japanese Women Between 1993 and 2008: A Panel Data Analysis. Australian National University, Asia Economic Papers No. 396,
https://crawford.anu.edu.au/pdf/pep/apep-396.pdf (24.11.2013).
Macnaughtan, Helen. 2006. 'From 'Post-War' to 'Post-Bubble': Contemporary Issues for Japanese Working Women.' In Perspectives on Work, Employment and Society in Japan, ed. Peter Matanle and Wim Lunsing, Basingstoke: Palgrave: 31–57.
MEXT (Ministry of Education, Culture, Sports, Science and Technology). 2013. Gakkō kihon chōsa, nenji tōkei, tōkatsuhyō 4 (School Basic Survey; Annual Statistics; Summary Tabulations 4). http://www.e- stat.go.jp/SG1/estat/List.do?bid=000001015843&cycode=0 (26.11.2013).
MHLW (Ministry of Health, Labour, and Welfare). 2012a. Rōdōryoku chōsa. Chōkiji keiretsu dēta. Hyō 9 Nenrei kaikyū koyō keitai betsu koyōsha sū (Labour Force Survey; Time Series Data. Table 9: Number of Employees by Age and Type of Employment).
http://www.stat.go.jp/data/roudou/longtime/03roudou.htm#hyo_9. Data file at:
http://www.stat.go.jp/data/roudou/longtime/zuhyou/lt51.xls (31.05.2013).
MHLW (Ministry of Health, Labour, and Welfare). 2012b. Chingin kōzō kihon tōkei chōsa (Basic Survey on Wage Structure). http://www.mhlw.go.jp/toukei/itiran/roudou/chingin/kouzou/z2012/; Data compilation provided by Japan Institute for Labour Policy Training (JILPT).
Mikanagi Yumiko. 2011. 'The Japanese Conception of Citizenship'. In The Routledge Handbook of Japanese Politics, ed. Alisa Gaunder, New York: Routledge: 130–139.
Miura Atsushi. 2007. Karyū shakai dai-ni-shō: Naze otoko wa onna ni “maketa” no ka (Low-Stream Society, Part Two: Why Are Men “Outplayed” By Women?). Tokyo: Kōbunsha.
Miyamoto Dai and Nakata Yoshifumi. 2002. 'Seiki jūgyōin no koyō sakugen to hisei rōdō no zōka: 1990 nendai no ōgata ko'urigyō o taishō ni' (The Decrease of Regular Employees and the Increase of Non-Regular Employees: The Case of Large Retailers in the 1990s). In Risutora to tenshoku no mekanizumu (The
Mechanism of Restructuring and Job Change), ed. Genda Yūshi and Nakata Yoshifumi. Tokyo: Tōyō Keizai Shinpōsha.
Nagase Nobuko. 2002. 'Jakunensō no koyō no hiseika to kekkon kōdō' (Marriage Timing and the Effect of Increase in Non-Standard Employment Among the Youth in Japan). Jinkō mondai kenkyū (Journal of Population Problems) 58 (2): 22–35.
OECD. 2009. 2009 OECD Employment Outlook. Tackling the Job Crisis. OECD Publishing.
Ōtake, Fumio. 2005. Nihon no fubyōdō: Kakusa shakai no gensō to mirai (Inequality in Japan - Illusion and Future of the Divided Society). Tokyo: Nikkei.
Ono Akiko. 2001. 'Ōgata ko'urigyō ni okeru bumon no gyōseki kanri to pātotaimā' (Earnings Management at Large Retailers and Part-time Employees). Nihon rōdō kenkyū zasshi (The Japanese Journal of Labour Studies) 498: 99–109.
Owen, George W. and Seppo Honkapohja. 2011. Learning and Expectations in Macroeconomics, Princeton:
Princeton University Press.
Satō Hiroki. 2001. ‘Atypical Employment: A Source of Flexible Work Opportunities?’ Social Science Japan Journal 4 (2): 161–181.
Shire, Karen A. 2008. 'Gender Dimension of the Ageing Workforce.' In The Demographic Challenge: A Handbook about Japan, ed. Florian Coulmas, Harald Conrad and Annette Schad-Seifert, Leiden: Brill Academic Publishing: 963–978.
Shire, Karen A. and Imai Jun. 2000. Flexible Equality: Men and Women in Employment in Japan.
Duisburger Arbeitspapiere Ostasienwissenschaften No. 30. http://ideas.repec.org/p/zbw/udedao/302000.html (02.10.2013).
Song, Jiyeoun. 2010. 'Japan's Labor Market Reform After the Collapse of the Bubble Economy: Political Determinants of Regulatory Changes'. Asian Survey 50 (6): 1011-1031.
Steinberg, Chad. 2012. Can Women Save Japan (and Asia Too)? International Monetary Fund Working Paper No. 12/248, http://www.imf.org/external/pubs/cat/longres.aspx?sk=40048.0 (24.11.2013).
Suzuki Munenori, Itō Midori, Ishida Mitsunori, Nihei Norihiro and Maruyama Masao. 2010. 'Individualizing Japan: Searching for its Origin in First Modernity.' The British Journal of Sociology 61: 513–538.
Tachibanaki, Toshiaki. 1998. Nihon no keizai kakusa: Shotoku to shisan kara kangaeru (Japan's Economic Disparities - Thoughts on Income and Assets). Tokyo: Iwanami Shinsho.
____ 2008. Jojo kakusa (Inequality Among Women). Tokyo: Tōyō Keizai Shinpōsha.
Weathers, Charles. 2001. ‘Changing White‐Collar Workplaces and Female Temporary Workers in Japan’.
Social Science Japan Journal 4(2): 201–218.
Works Institute. 2012. 'Dai 29 kai wākusu daisotsu kyūjin bairitsu chōsa' (The 29th Works-Survey On Job- To-Applicant Ratios For University Graduates).
http://www.recruit.jp/news_data/library/pdf/20120425_01.pdf (30.05.2013).
Yamada Kazuyo. 2009. ‘Past and Present Constraints on Labor Movements for Gender Equality in Japan’.
Social Science Japan Journal 12 (2): 195–209.
Yamada Ken. 2011. 'Labour Supply Responses to the 1990s Japanese Tax Reforms'. Labour Economics 18:
539–546.
Yamagami Toshihiko. 2006. 'Shuntō no genkai bea fukkatsu ga “kakusa shakai” o jochō suru' (Limits to the Union Power: Where Returning Bear Markets Foster Societal Gaps), Shūkan ekonomisuto (Weekly
Economist) 28 February: 84–87.
Yamaguchi Masao. 2011. 'Sei'shain to hisei'shain no daitai hokan kankei ni kansuru keiryō bunseki' (An Empirical Analysis of the Substitution and Complementarity Between Regular and Non-Regular Employees in Japanese Firms). Nihon keizai kenkyū (Studies on the Japanese Economy) 64: 27–55.
Annex 1: Share of standard employment by age cohorts, 1988-2009
Source: Own calculations based on data from the Statistics Bureau Japan, Historical Statistics, Chapter 2 (Population
